{"id":"f6fc5831-d870-414c-a6a3-1dfc05a429c6","arxiv_id":"2607.11476","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":1,"one_line_summary":"DarkSide-50 sets the first experimental lower limit on the two-neutrino double electron capture half-life of 36Ar at 9.2×10^19 years (90% C.L.).","lead":"No excess of events was found in a search for two-neutrino double electron capture in argon-36 using the DarkSide-50 detector, yielding a lower half-life limit of 9.2×10^19 years at 90% confidence. This is the first experimental constraint on this decay mode and sets a benchmark for future argon-based experiments.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Total-energy signal model validated with a poor χ² (87.75/59, p≈0.008) yet described as 'well within'; ±2 Ne shift only tests energy offset, not shape distortions, so the 36Ar limit may be overconfident.","rationale":"The reader correctly identified the signal-modelling assumption as the weakest point, but treated the ±2 Ne systematic shift as sufficient. My read is that the shift does not cover shape uncertainties, and the validation fit is actually poor (χ²/ndf = 1.49, p ≈ 0.008), which is not acknowledged in the text. Because the reported T1/2 limit is the paper's central claim, a biased signal template could directly bias that limit. The proposed conditional change is not an outright rejection: the paper's logic is sound and the analysis is reproducible in principle, but the robustness check I propose would settle whether the limit is stable under realistic spectral variations. If the limit does not change significantly, the ACCEPT verdict stands; if it does change, the paper needs a more conservative treatment. Hence CONDITIONAL is the appropriate verdict adjustment.","tokens_in":14472,"tokens_out":17893,"duration_ms":161199,"concrete_test":"Recompute the 90% CL upper limit on N_2EC2ν using signal templates that (a) broaden both KK and KL peaks by 50% (σ → 1.5σ), (b) apply a common energy-scale shift of ±5 Ne, and (c) use the independent-electron cascade model (which failed the 37Ar fit but represents the opposite extreme). Track how the resulting T1/2 limit changes relative to the nominal 9.2×10^19 yr; if the variation exceeds 20%, the signal-shape systematic is underestimated and the stated limit is not robust. Additionally, refit the 37Ar spectrum with a free energy-scale nuisance parameter and report the best-fit shift; if that best-fit shift exceeds ±2 Ne, the adopted systematic is unjustified.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim is the reported half-life limit T1/2 > 9.2×10^19 yr, which relies critically on the simulated energy spectrum of the 36Ar 2EC2ν signal. The signal model is the 'total-energy approach' validated against 37Ar K-shell data in Sec. 6. However, the fit to the 37Ar spectrum yields χ²/ndf = 87.75/59, corresponding to p ≈ 0.008 — a statistically significant discrepancy that contradicts the text's assertion that residuals are 'well within the statistical uncertainties'. The systematic correction introduced is a simple ±2 Ne shift of the entire signal spectrum. This tests only an overall energy-scale offset; it does not test shape distortions such as a broadening of the double-vacancy cascade response or a different recombination behaviour due to overlapping ionization clouds, which the paper explicitly notes is a concern. If the true 36Ar signal is broader or has a different peak structure, the profile-likelihood test statistic under the signal hypothesis would be reduced, potentially leading to an overestimated 90% CL upper limit on N_2EC2ν and hence an overestimated T1/2 limit. Since this is the first constraint on 2EC2ν in 36Ar, the validity of the headline limit rests on this unvalidated shape assumption. The paper's claim that the ±2 Ne shift does not change the result addresses only one narrow systematic axis.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports the first experimental search for two-neutrino double electron capture (2EC2ν) in 36Ar using the DarkSide-50 liquid argon detector. With a 12.3 ton-day underground argon exposure, no statistically significant excess is observed, and a 90% C.L. lower limit on the half-life is set at T1/2 > 9.2×10^19 yr. The signal is modeled by collapsing the full atomic de-excitation cascade into a single monoenergetic energy deposit, validated against 37Ar calibration data. The analysis uses a binned profile likelihood with background components from 39Ar, 85Kr, PMTs, and the cryostat, and the 36Ar isotopic abundance is measured via ICP-MS. A projected sensitivity for DarkSide-20k is also discussed.","tokens_in":14785,"tokens_out":6282,"duration_ms":56728,"significance":"If accepted, this paper provides the first experimental constraint on 2EC2ν in 36Ar, a decay whose predicted half-life (~1.7×10^29 yr) is far beyond current sensitivity, making the limit a first step rather than a test of nuclear matrix element calculations. The analysis benefits from a well-established detector, a previously validated background model, and a standard profile likelihood treatment. The external measurement of the 36Ar abundance and the transparent conversion to a half-life limit are strengths. The main risk is the signal model: the total-energy approximation is an acknowledged simplification, and its validation against 37Ar shows a statistically poor fit, which could affect the central limit.","major_comments":[{"comment":"The validation fit of the total-energy model to the 37Ar spectrum reports χ²/ndf = 87.75/59. With 59 degrees of freedom, this corresponds to p ≈ 0.008, i.e. a statistically significant discrepancy, rather than 'well within the statistical uncertainties' as stated in the text. Since the 36Ar signal model relies on the same total-energy approximation, this poor goodness-of-fit is directly relevant to the central limit. The ±2 Ne systematic shift tests only an overall energy-scale offset; it does not cover shape distortions (e.g., broadening or altered recombination in the double-vacancy cascade). I request a quantitative treatment of signal-shape uncertainty (e.g., an additional broadening parameter or a conservative alternative shape) and its propagation to the final half-life limit.","section":"Sec. 6, Fig. 3"},{"comment":"The half-life limit is obtained from the fitted signal-strength limit multiplied by the signal model s_i. If the true 36Ar cascade response is broader than the monoenergetic total-energy model, the profile-likelihood ratio for a given signal strength changes, and the 90% C.L. bound on N_2EC2ν (278 events) could move. Since this is the first constraint on 2EC2ν in 36Ar, the validity of the headline T1/2 limit rests on the signal-shape assumption. The statement that a ±2 Ne shift leaves the limit unchanged addresses only one systematic axis. Please demonstrate robustness to shape variations or incorporate them as nuisance parameters.","section":"Sec. 8, Eq. (6)"}],"minor_comments":[{"comment":"Typo: 'we we employ a similar approach' should read 'we employ a similar approach'.","section":"Sec. 6"},{"comment":"Typo: 'the active mass ... is givern by' should read 'is given by'.","section":"Sec. 8"},{"comment":"The projected DarkSide-20k sensitivity is stated inconsistently: the abstract says a factor ~100, Sec. 9 says 'two to three orders of magnitude', and Sec. 10 says 'about two orders of magnitude'. Please harmonize.","section":"Abstract / Sec. 9 / Sec. 10"},{"comment":"The legend text appears garbled (' yr) 1910×>9.21/22EC2v (TPMTs'); the formatting should be corrected for clarity.","section":"Fig. 5"},{"comment":"Minor grammatical issue: 'offering a unique pathway to determination the absolute neutrino mass scale' should read 'to determining the absolute neutrino mass scale'.","section":"Sec. 1"}],"recommendation":"major_revision","confidential_remarks":"The core analysis is sound and the first limit is valuable, but the 37Ar validation p-value (≈0.008) contradicts the text's characterization and is load-bearing for the signal-shape assumption. This is fixable with a quantitative shape systematic, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe short version: this is a legitimate first limit, not a breakthrough. DarkSide-50 has searched for two-neutrino double electron capture in 36Ar and set T1/2 > 9.2e19 yr at 90% CL. No prior constraint existed, so that's a genuinely new data point. The strongest part is that the analysis piggybacks on the well-established DS-50 low-mass WIMP framework—the background model, selection cuts, and profile likelihood have already been vetted in earlier publications. The 36Ar abundance in UAr is not assumed from air; it is measured with ICP-MS, giving a depletion factor of 45.6±0.8. That's real work.\n\nThe weak spot is the signal model validation—and the paper somewhat overstates it. The total-energy approach (collapse the entire cascade to a single monoenergetic deposit) is an approximation. It is tested against the 37Ar K-capture line, and the fit gives χ²/ndf=87.75/59, which corresponds to p≈0.008. Calling that 'well within the statistical uncertainties' is not defensible. The ±2 Ne shift only moves the peak position; it does not test shape distortions, such as a broadening of the response from the overlapping double-vacancy clouds. If the true 36Ar signal is broader than modeled, the limit would weaken. That concern is real but probably not fatal: the predicted half-life is ~1.7e29 yr, ten orders of magnitude away, and the headline result is a null observation. Still, a referee should ask for a clearer statement about the χ² and, if feasible, a shape-variation systematic.\n\nThe DarkSide-20k projection is a scaling estimate, not a full sensitivity projection; treat it as a teaser.\n\nOverall: it's a carefully done, modest result. It deserves a serious referee; the signal-model validation needs to be addressed in revision. I'd probably cite it if I worked in the area, and I think the atomic-cascade discussion makes it a decent reading-group topic.","headline":"First limit on 2EC2ν in 36Ar from DarkSide-50: solid null result, but the signal-model validation is weaker than claimed.","tokens_in":15734,"tokens_out":2597,"would_cite":true,"duration_ms":23949,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper reports the first experimental search for two-neutrino double electron capture in 36Ar and sets a lower limit of 9.2×10^19 years on its half-life at 90% confidence.","keywords":["double electron capture","36Ar","DarkSide-50","liquid argon TPC","half-life limit","rare nuclear decay","underground argon","profile likelihood"],"falsifier":"A concrete falsifier would be a measurement of the 36Ar double-vacancy de-excitation spectrum that shows the energy deposition is not monoenergetic in the way assumed—for example, an atomic-physics measurement of the KK and KL cascade in 36S that predicts a significantly different ionization response distribution (e.g., a peak shift larger than the ±2 electron systematic) in liquid argon. Alternatively, a reanalysis of the same DarkSide-50 data using a first-principles cascade simulation, if it produced a statistically significant excess in the 25–90 electron ROI, would directly contradict the","tokens_in":14371,"feed_emoji":"⚛️","tokens_out":4925,"duration_ms":38703,"temperature":0.7,"pith_summary":"This paper reports the first search for two-neutrino double electron capture (2EC2ν) in 36Ar, using the underground-argon data from the DarkSide-50 detector. The process would leave a two-vacancy atomic de-excitation cascade of about 4.8 keV total energy, which the authors model as a monoenergetic energy deposit. No statistically significant excess is observed above background in the ~12 ton-day exposure, and the authors translate this null result into a lower limit on the half-life of 9.2×10^19 years at 90% confidence. They also project that the upcoming DarkSide-20k experiment could extend sensitivity by about two orders of magnitude. This matters because 36Ar is one of only a dozen isotopes where double electron capture is the sole allowed double-beta channel, and the measurement provides a new experimental anchor for nuclear matrix element calculations.","feed_headline":"First limit on 36Ar double electron capture: >9e19 yr","feed_subtitle":"A 12-ton-day underground argon search finds no signal, ruling out half-lives below 9.2×10^19 years.","key_machinery":"The central object is the total-energy approximation for the atomic de-excitation cascade: the entire energy release from the double-vacancy cascade (~4.9 keV for KK capture and ~2.6 keV for KL capture) is collapsed into a single monoenergetic interaction point in the detector response model. This approximation is validated against the 37Ar electron-capture K and L lines and is carried over to 36Ar. The second key ingredient is the measured depletion factor of 36Ar in underground argon (45.6±0.8 relative to atmospheric argon), which enters the half-life formula. The profile likelihood ratio with asymptotic formulae converts the observed event count into an upper limit.","core_discovery":"The central claim is that the two-neutrino double electron capture of 36Ar has not been observed, and the first experimental lower limit on its half-life is T_{1/2} > 9.2×10^19 yr (90% C.L.). The result comes from a binned profile likelihood analysis of 633.5 live-days of DarkSide-50 data with a 19.4 kg fiducial liquid argon target, using underground argon whose 36Ar isotopic abundance was measured to be (7.32±0.13)×10^-3%. The signal model treats the full energy release of the KK and KL capture de-excitation cascades as a single monoenergetic deposit, a total-energy approach validated against 37Ar electron-capture calibration data. No signal events are found; the best-fit number of signal e","pith_inferences":["If the monoenergetic total-energy approximation fails for overlapping double-vacancy clouds in a way not covered by the ±2 electron systematic shift, the true signal could be broader or shifted, and the limit could be biased; a dedicated atomic-cascade measurement (e.g., with a trapped 36Ar source or a high-resolution microcalorimeter) could test this.","The measured 36Ar depletion factor in underground argon means other underground argon experiments (like DarkSide-20k and ARGO) inherit the same abundance; if the depletion varies by source, their sensitivity projections may need adjustment.","An independent measurement of the 36Ar 2EC2ν half-life using a different technique, such as a bolometric detector or a gaseous TPC with enriched 36Ar, would provide a cross-check of the dark-matter detector-based result.","Should theoretical NME calculations improve and predict a half-life closer to current sensitivity, the same dataset could be re-analyzed with a reoptimized ROI, potentially yielding a stronger limit or a hint."],"forward_implications":["If the limit is correct, any claimed observation of 36Ar 2EC2ν in future experiments would have to involve half-lives longer than 9.2e19 yr, or point to new physics or a nuclear-model breakdown.","The projected sensitivity of DarkSide-20k (~1e22 yr) could either discover the decay or push the constraint closer to the theoretical prediction of 1.7e29 yr, helping to test the USD shell-model matrix elements.","The total-energy modeling validated on 37Ar provides a template for low-energy rare-decay searches in liquid argon TPCs.","The null result adds a new data point to the set of double electron capture half-life limits, complementing measurements in 124Xe and 78Kr."],"fun_headline_variants":["No 36Ar double electron capture found: half-life > 9.2e19 yr","DarkSide-50 sets first bound on 36Ar double electron capture","36Ar double electron capture: null result yields >9e19 yr limit","Rare argon decay ruled out: T1/2 > 9.2e19 yr","Underground argon search: no 36Ar double electron capture"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The signal model assumes that the entire double-vacancy de-excitation cascade of 36Ar can be treated as a single monoenergetic energy deposit, an approximation the paper itself notes is not feasible to derive from first principles; if this total-energy response is wrong for the overlapping double-vacancy electron clouds, the extracted half-life limit could change.","fun_headline_variants_meta":{"raw":{"variants":["No 36Ar double electron capture found: half-life > 9.2e19 yr","DarkSide-50 sets first bound on 36Ar double electron capture","36Ar double electron capture: null result yields >9e19 yr limit","Rare argon decay ruled out: T1/2 > 9.2e19 yr","Underground argon search: no 36Ar double electron capture"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000929,"raw_usage":{"total_tokens":3837,"prompt_tokens":790,"completion_tokens":3047,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":534,"completion_tokens_details":{"reasoning_tokens":2941}},"tokens_in":534,"tokens_out":3047,"duration_ms":19050,"temperature":1.0,"reasoning_tokens":2941,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T06:53:20.694828+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete falsifier would be a measurement of the 36Ar double-vacancy de-excitation spectrum that shows the energy deposition is not monoenergetic in the way assumed—for example, an atomic-physics measurement of the KK and KL cascade in 36S that predicts a significantly different ionization response distribution (e.g., a peak shift larger than the ±2 electron systematic) in liquid argon. Alternatively, a reanalysis of the same DarkSide-50 data using a first-principles cascade simulation, if it produced a statistically significant excess in the 25–90 electron ROI, would directly contradict the","supporting_citations":[],"review_version":2}